By MIKE STO­BBE-As­so­ci­at­ed Press

NEW YORK (AP) — As coro­n­avirus in­fec­tions rise in some parts of the world, ex­perts are watch­ing for a po­ten­tial new COVID-19 surge in the U.S. — and won­der­ing how long it will take to de­tect.

De­spite dis­ease mon­i­tor­ing im­prove­ments over the last two years, they say, some re­cent de­vel­op­ments don’t bode well:

—As more peo­ple take rapid COVID-19 tests at home, few­er peo­ple are get­ting the gold-stan­dard tests that the gov­ern­ment re­lies on for case counts.

—The Cen­ters for Dis­ease Con­trol and Pre­ven­tion will soon use few­er labs to look for new vari­ants.

—Health of­fi­cials are in­creas­ing­ly fo­cus­ing on hos­pi­tal ad­mis­sions, which rise on­ly af­ter a surge has ar­rived.

—A waste­water sur­veil­lance pro­gram re­mains a patch­work that can­not yet be count­ed on for the da­ta need­ed to un­der­stand com­ing surges.

White House of­fi­cials say the gov­ern­ment is run­ning out of funds for vac­cines, treat­ments and test­ing.

“We’re not in a great sit­u­a­tion,” said Jen­nifer Nuz­zo, a Brown Uni­ver­si­ty pan­dem­ic re­searcher.

Sci­en­tists ac­knowl­edge that the wide avail­abil­i­ty of vac­cines and treat­ments puts the na­tion in a bet­ter place than when the pan­dem­ic be­gan, and that mon­i­tor­ing has come a long way.

For ex­am­ple, sci­en­tists this week tout­ed a 6-month-old pro­gram that tests in­ter­na­tion­al trav­el­ers fly­ing in­to four U.S. air­ports. Ge­net­ic test­ing of a sam­ple on Dec. 14 turned up a coro­n­avirus vari­ant — the de­scen­dant of omi­cron known as BA.2 — sev­en days ear­li­er than any oth­er re­port­ed de­tec­tion in the U.S.

More good news: U.S. cas­es, hos­pi­tal­iza­tions and deaths have been falling for weeks.

But it’s dif­fer­ent else­where. The World Health Or­ga­ni­za­tion this week re­port­ed that the num­ber of new coro­n­avirus cas­es in­creased two weeks in a row glob­al­ly, like­ly be­cause COVID-19 pre­ven­tion mea­sures have been halt­ed in nu­mer­ous coun­tries and be­cause BA.2 spreads more eas­i­ly.

Some pub­lic health ex­perts aren’t cer­tain what that means for the U.S.

BA.2 ac­counts for a grow­ing share of U.S. cas­es, the CDC said — more than one-third na­tion­al­ly and more than half in the North­east. Small in­creas­es in over­all case rates have been not­ed in New York, and in hos­pi­tal ad­mis­sions in New Eng­land.

Some of the north­ern U.S. states with the high­est rates of BA.2, how­ev­er, have some of the low­est case rates, not­ed Ka­tri­ona Shea of Penn State Uni­ver­si­ty.

Dr. James Muss­er, an in­fec­tious dis­ease spe­cial­ist at Hous­ton Methodist, called the na­tion­al case da­ta on BA.2 “murky.” He added: “What we re­al­ly need is as much re­al-time da­ta as pos­si­ble ... to in­form de­ci­sions.”

Here’s what COVID-19 track­ers are look­ing at and what wor­ries sci­en­tists about them.

TEST RE­SULTS

Tal­lies of test re­sults have been at the core of un­der­stand­ing coro­n­avirus spread from the start, but they have al­ways been flawed.

Ini­tial­ly, on­ly sick peo­ple got test­ed, mean­ing case counts missed peo­ple who had no symp­toms or were un­able to get swabbed.

Home test kits be­came wide­ly avail­able last year, and de­mand took off when the omi­cron wave hit. But many peo­ple who take home tests don’t re­port re­sults to any­one. Nor do health agen­cies at­tempt to gath­er them.

Mara As­pinall is man­ag­ing di­rec­tor of an Ari­zona-based con­sult­ing com­pa­ny that tracks COVID-19 test­ing trends. She es­ti­mates that in Jan­u­ary and Feb­ru­ary, about 8 mil­lion to 9 mil­lion rapid home tests were be­ing done each day on av­er­age — four to six times the num­ber of PCR tests.

Nuz­zo said: “The case num­bers are not as much a re­flec­tion of re­al­i­ty as they once were.”

HUNT­ING FOR VARI­ANTS

In ear­ly 2021, the U.S. was far be­hind oth­er coun­tries in us­ing ge­net­ic tests to look for wor­ri­some virus mu­ta­tions.

A year ago, the agency signed deals with 10 large labs to do that ge­nom­ic se­quenc­ing. The CDC will be re­duc­ing that pro­gram to three labs over the next two months.

The week­ly vol­ume of se­quences per­formed through the con­tracts was much high­er dur­ing the omi­cron wave in De­cem­ber and Jan­u­ary, when more peo­ple were get­ting test­ed, and al­ready has fall­en to about 35,000. By late spring, it will be down to 10,000, al­though CDC of­fi­cials say the con­tracts al­low the vol­ume to in­crease to more than 20,000 if nec­es­sary.

The agency al­so says turn­around time and qual­i­ty stan­dards have been im­proved in the new con­tracts, and that it does not ex­pect the change will hurt its abil­i­ty to find new vari­ants.

Out­side ex­perts ex­pressed con­cern.

“It’s re­al­ly quite a sub­stan­tial re­duc­tion in our base­line sur­veil­lance and in­tel­li­gence sys­tem for track­ing what’s out there,” said Bron­wyn MacIn­nis, di­rec­tor of pathogen ge­nom­ic sur­veil­lance at the Broad In­sti­tute of MIT and Har­vard.

SEWAGE SUR­VEIL­LANCE

An evolv­ing mon­i­tor­ing sys­tem is look­ing for signs of coro­n­avirus in sewage, which could po­ten­tial­ly cap­ture brew­ing in­fec­tions.

Re­searchers have linked waste­water sam­ples to the num­ber of pos­i­tive COVID-19 tests a week lat­er, sug­gest­ing health of­fi­cials could get an ear­ly glimpse at in­fec­tion trends.

Some health de­part­ments al­so have used sewage to look for vari­ants. New York City, for ex­am­ple, de­tect­ed sig­nals of the omi­cron vari­ant in a sam­ple tak­en on Nov. 21 — about 10 days be­fore the first case was re­port­ed in the U.S.

But ex­perts note the sys­tem doesn’t cov­er the en­tire coun­try. It al­so doesn’t dis­tin­guish who is in­fect­ed.

“It’s a re­al­ly im­por­tant and promis­ing strat­e­gy, no doubt. But the ul­ti­mate val­ue is still prob­a­bly yet to be un­der­stood,” said Dr. Jeff Duchin, the health of­fi­cer for Seat­tle/King Coun­ty, Wash­ing­ton.

HOS­PI­TAL DA­TA

Last month, the CDC out­lined a new set of mea­sures for de­cid­ing whether to lift mask-wear­ing rules, fo­cus­ing less on pos­i­tive test re­sults and more on hos­pi­tals.

Hos­pi­tal ad­mis­sions are a lag­ging in­di­ca­tor, giv­en that a week or more can pass be­tween in­fec­tion and hos­pi­tal­iza­tion. But a num­ber of re­searchers be­lieve the change is ap­pro­pri­ate. They say hos­pi­tal da­ta is more re­li­able and more eas­i­ly in­ter­pret­ed than case counts.

The lag al­so is not as long as one might think. Some stud­ies have sug­gest­ed many peo­ple wait to get test­ed. And when they fi­nal­ly do, the re­sults aren’t al­ways im­me­di­ate.

Spencer Fox, a Uni­ver­si­ty of Texas da­ta sci­en­tist who is part of a group that us­es hos­pi­tal and cell­phone da­ta to fore­cast COVID-19 for Austin, said “hos­pi­tal ad­mis­sions were the bet­ter sig­nal” for a surge than test re­sults.

There are con­cerns, how­ev­er, about fu­ture hos­pi­tal da­ta.

If the fed­er­al gov­ern­ment lifts its pub­lic health emer­gency de­c­la­ra­tion, of­fi­cials will lose the abil­i­ty to com­pel hos­pi­tals to re­port COVID-19 da­ta, a group of for­mer CDC di­rec­tors re­cent­ly wrote. They urged Con­gress to pass a law that will pro­vide en­dur­ing au­thor­i­ties “so we will not risk fly­ing blind as health threats emerge.”

___

AP re­porters Lau­ran Neer­gaard in Wash­ing­ton and Lau­ra Un­gar in Louisville, Ken­tucky, con­tributed.